Objectives: The objective of this K24 proposal is to enable the candidate to mentor young investigators to successfully integrate clinical, outcomes and health services research methods to optimize management strategies for gastrointestinal disease. Background: The candidate has a successful program of patientoriented research in gastrointestinal disease. He has obtained independent NIH funding for his research, and has an established record of mentoring young investigators who have been successful in conducting patientoriented research, publishing in peer-reviewed journals and obtaining NIH support in the form of K-awards. The candidate has been invited to teach patient-oriented research design to national and international audiences, and publish reviews of research techniques in leading journals. Methods: Mentoring plan: The candidate's environment allows new investigators to obtain formal training in clinical study design, epidemiology and biostatistics. Specific goals for the candidate include formal coursework in health economic theory and participation in the UCSF Clinical and Translational Science Institute Mentor Development Program. The candidate proposes to teach quantitative analysis in order to generate hypotheses, which can be tested through properly designed prospective clinical trials. Data derived from these studies can be integrated into revised decision models to identify the optimal competing strategies of management, which can then be prospectively tested. An example testing this hypothesis is presented as a research project, which uses the infrastructure established by the candidate's R01 project to recruit and follow subjects. Research plan: The effectiveness of strategies to screen for colorectal neoplasia depends heavily on the rate of adherence between screening tests. This study aims to identify factors associated with non-adherence to colorectal neoplasia screening in a vulnerable population at high-risk for poor health outcomes. We are conducting a prospective cohort study of patients at average risk for development of polyps in which subjects receive counseling about screening using fecal occult blood testing or colonoscopy or their choice of either test. This proposal adds a survey to identify factors based on the Health Belief Model, including sociodemographic variables that predict non-adherence to screening. Data Collection: 1. The primary outcome is identification of variables associated with non-adherence. Data Analysis: Chi-square testing of proportions and generalized linear models. Public health: This grant will support mentorship of new investigators interested in patient-oriented research in gastrointestinal disease. We aim to identify important questions through quantitative analysis and then definitively answer these questions through prospective clinical trials. The research project will provide targets for intervention to increase screening adherence for pre-malignant colorectal neoplasia, focusing on vulnerable populations.